3. Terrain Reconstruction: Map Regression and Military Terrain Analysis

The ease with which Dussindale's location can be discerned greatly enhances the site's suitability for terrain reconstruction and analysis. Map regression has been employed at American battlefields since the early 1980s (Scott and McFeaters 2011) and has increasingly been adopted in Britain since the mid-1990s, where it has formed a vital tool in the exploration of engagements including Edgehill, Sedgemoor, and Bosworth (Foard 2007; 2008; 2013). This process begins by 'registering' a suitable base map to a modern Ordnance Survey (OS) map, ensuring that terrain features and buildings recorded on the former appear in the same position on the latter (Foard and Morris 2012). In the case of English battlefields, first edition six-inch-to-one-mile OS map sheets dating from the 19th century are the best available option, retaining the accuracy of their later counterparts but being drafted before the major industrialisation and development of the modern era (Foard 2008). Once the base map has been 'registered', progressively earlier representations of the area can be considered in the same fashion, matching each source to surviving terrain features and tracing it onto the image at an appropriate scale. The use of GIS such as MapInfo considerably simplifies this stage by allowing maps to be traced into a number of editable 'layers', which can be superimposed and switched off as required.

Notwithstanding the precision of such digital tools, map regression is far from an exact science, and involves an incremental decision-making process wherein the information acquired from different sources is consolidated within a single composite image, rather than representing a direct reproduction of the earliest map. This issue is particularly pronounced when working with older works, such as the late 16th-century maps of Mousehold Heath, which are often prone to errors including inconsistent internal scales, variable production quality, and insufficient accuracy. In the case of the 1589 map (Figure 3), for instance, this is manifested in the lack of a stated scale of measurement and the abstract representation of settlements, giving little indication of the distances involved, or of the extent of urban areas. Similarly, the c. 1600 map (Figure 4) is almost certainly a later, reduced-scale reproduction, compressing some but not all of the same topographical features onto a far smaller image, with resultant implications for the accuracy of the representation. Documentary sources, provided by the glebe terriers of church lands, the 1812 enclosure award, and, most crucially, Ward and Harrison's texts, offer a means of assistance where maps are unavailable, unclear, or omit vital information, enabling such elisions to be redressed. Nonetheless, the resultant composite map will always contain an element of uncertainty because of potential unknown features or unrecorded changes within the landscape, and so cannot be regarded as completely definitive. Despite these issues, map regression fulfils a critical role in conflict studies by facilitating the investigation of an engagement in the context of its historic, rather than modern, terrain, allowing scholars to re-evaluate battlefields where the area's original landscape has altered over the intervening centuries.

The successful application of these techniques is a vital prerequisite to the process of military terrain analysis, which seeks to surmise the probable deployment and tactical movement of opposing armies in situations where narrative sources are sparse, contradictory, or elliptical. This approach has evolved from Burne's (1950) theory of Inherent Military Probability, which suggested that army commanders would deploy their forces according to universal requirements, into a more nuanced concept, known by the acronyms KOCOA, OAKOC or OCOKA, derived from modern tactical guidelines (Reserve Officer Training Corps (ROTC) 2003). Where Burne (1950) interpreted a battlefield as a whole, assessing how a trained soldier would seek to exploit the landscape, the KOCOA method breaks this process down into a series of quantifiable stages, identified as Key Terrain, Observation and Fields of Fire, Cover and Concealment, Obstacles, and Avenues of Approach and Withdrawal. The first category, key terrain, defines topographical features that exert an impact during combat, while those of observation and fields of fire and cover and concealment determine where soldiers can most effectively be deployed to attain or deny visibility to their enemies (ROTC 2003). Obstacles encompass pre-existing or man-made impediments that can be exploited or created to disrupt, redirect, or block enemy movement, and the final classification, avenues of approach and withdrawal, represents both the routes by which armies enter and leave the battlefield, and the access points to terrain features within the combat zone (ROTC 2003).

The use of these criteria can compensate for the often limited information available regarding tactical deployment during pre-modern engagements, but only if the era's common battlefield doctrines, not to mention the forces utilised in the action in question, can be accurately defined (Scott and McFeaters 2011). Without such period-specific detail, scholars risk the anachronistic application of current military principles to historic battles, irrespective of the crucial differences in tactical thought and technology between modern and late medieval warfare (Carman and Carman 2006). In short, successful military terrain analysis rests upon an understanding of how armies from the period would typically approach the tactical problems and opportunities offered by a particular landscape, a caveat referred to as Inherent Historic Military Probability (Foard 2008). For instance, where a modern commander might place infantrymen in small, dispersed units amidst woods or urban terrain, to provide concealment and protection from incoming fire, his medieval or Renaissance antecedent would be more concerned with finding open ground on which to array large, densely packed formations containing thousands of soldiers. Providing, then, that due attention is paid to the period's standard tactical deployments, as far as they can be discerned from military manuals and other engagements, military terrain analysis can aid in the interpretation of battlefield positioning within a reconstructed landscape. This has been demonstrated by numerous examples within US battlefield studies, for instance at Chelsea Creek (Maio et al. 2013), Gettysburg (National Park Service (NPS) 2014), and Buckland Mills (NPS 2008), and is equally applicable to Dussindale, where informed inferences can be made to link narratives of the action with its historic landscape.